iLiad The six unused connections

[design256]

I have finally had a chance to look at the spare connections on the Iliad. This is what I found:

13 - wired to 17 in the plug, but held to ground when disconnected
14 - High impedance, either an input or N/C
15 - signal on this line (see below)
16 - held to ground
17 - wired to 13, high impedance when disconnected, probably plug sense input
18 - High impedance, either an input or N/C


The only interesting pin was 15. This is nominally high (3.5v), but emits a start bit followed series of 16 data bits at approximately 9600 baud whenever a key is pressed or released. As far as I can tell, no other data (like a console) is present on this pin.


Sometime soon I'll take the thing apart and have a poke around inside for anything that might be a console.

[Alexander Turcic]

Originally Posted by design256

Sometime soon I'll take the thing apart and have a poke around inside for anything that might be a console.

Cool. Just don't break anything ;-) And please keep us updated!

-> moved to the development section.

[arivero]

Originally Posted by design256

The only interesting pin was 15. This is nominally high (3.5v), but emits a start bit followed series of 16 data bits at approximately 9600 baud whenever a key is pressed or released. As far as I can tell, no other data (like a console) is present on this pin.

Well, it could be some sort of custom debug, some JTAG, or some real console with output debug at every keypress. One could try to attach a term to pins 14 and 15, but what about the flow control?

On other hand, there is still the internal connector if they want to have a serious debug.

[scotty1024]

RS-232 format is:
1 Start bit
8 Data bits
1 parity bit
1 or 2 stop bits (usually 1)

So at most 12 bits, not 16.

The only thing I can think of that is framed 16 bits and would be emitted when a "key is pressed" would be Tribble's much beloved "Thonk" soud e.g. a digital audio output pin.

But a single 16 bit frame wouldn't carry much Thonk in it. But they may not have gotten around to making it work correctly yet.

[DHer]

That's interesting.

I will try to think about it. At the moment i'd bet on custom debug.

[ali]

I just got my Iliad back. It's been opened, just to do a reinstall.

My files are still there, so I don't know what they did. But I think they wouldn't open it if there was a JTAG interface or any other write access in the external connector.

[tribble]

Thats sad.

[arivero]

Ah, can we asume you have captured the signal with a digital oscilloscope so you can guess the baud rate from the size of the pulses? Or did you use a different setup?

Also, where you able to count how long the whole series was?

Originally Posted by design256

I have finally had a chance to look at the spare connections on the Iliad. This is what I found:

13 - wired to 17 in the plug, but held to ground when disconnected
14 - High impedance, either an input or N/C
15 - signal on this line (see below)
16 - held to ground
17 - wired to 13, high impedance when disconnected, probably plug sense input
18 - High impedance, either an input or N/C


The only interesting pin was 15. This is nominally high (3.5v), but emits a start bit followed series of 16 data bits at approximately 9600 baud whenever a key is pressed or released. As far as I can tell, no other data (like a console) is present on this pin.


Sometime soon I'll take the thing apart and have a poke around inside for anything that might be a console.

[design256]

Originally Posted by arivero

Ah, can we asume you have captured the signal with a digital oscilloscope so you can guess the baud rate from the size of the pulses? Or did you use a different setup?

Also, where you able to count how long the whole series was?

Yes - the pulses were 100 microseconds minimum, giving approx 9600 baud, and each pulse train was 16 bits long. Further investigation has shown that the same pulse train seems to be sent whatever key is pressed.

[ali]

Originally Posted by design256

Further investigation has shown that the same pulse train seems to be sent whatever key is pressed.

Could it be much more than 9600 baud, so that the signal is not really same?

Whenever you press a key, something like this appears in the log, hence on the console:

<event time="012" type="button"><key>newspapers<key></event>

Could that be your signal?

[arivero]

Originally Posted by ali

Could it be much more than 9600 baud, so that the signal is not really same?
?

I doubt, any student machine capturates up to 10 Mhz without problem, and it seens unlikely to have a console running beyond 54Kbits. It could be a signal to the cradle to awake. Or it could be a software handshake for communication, with a counterpart to be expected in the input. A 16 bits Xon/Xoff.